amdrl

Asynchronous methods for deep reinforcement learning based on Distributed TensorFlow

Currently under construction! To try out one step Sarsa with CartPole-v0 try

$ ./launch_cluster --worker_tasks 4

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
agents		agents
configs		configs
models		models
net_fns		net_fns
transforms		transforms
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
environments.py		environments.py
launch_cluster.py		launch_cluster.py
memories.py		memories.py
parameter_server.py		parameter_server.py
policies.py		policies.py
runners.py		runners.py
worker.py		worker.py

Provide feedback